After weeks of anticipation, Microsoft confirmed on Monday that it is planning a multi-year, multi-billion dollar investment in OpenAI, the startup behind ChatGPT and Dall-E 2. The move puts pressure on the search engine giant, Google.
Microsoft initially partnered with the research institute in 2019, when it invested $1 billion. It then secured exclusive rights to OpenAI’s GPT-3 core language model in 2021. The New York Times reported that Microsoft’s new investment, revealed days after the tech giant laid off 10,000 employees , will be $10 billion.
Microsoft said that with this partnership, it will intensify the development and deployment of specialized supercomputer systems to accelerate the R&D of OpenAI which, for now, remains independent.
The vendor also plans to infuse the startup’s new models into its consumer and enterprise products. In the meantime, he launched Azure OpenAI services which allow companies to use the technologies of the startup to enrich their applications. Finally, Microsoft plans to power all OpenAI workloads across API research, products, and services.
Since OpenAI introduced ChatGPT, an NLG-based application capable of writing academic essays, code, answering massive numbers of questions with seemingly knowledgeable answers, in November, the system’s popularity around the world integer has soared, although concerns have been raised about its accuracy and ability to encourage plagiarism.
Microsoft is committed to AI
Despite reports from The New York Times and other outlets citing the $10 billion figure, a Microsoft spokesperson did not confirm the information with Techtarget. [propriétaire du MagIT]. “We are not disclosing financial details and terms of the agreement and we are not sharing other details about this extended partnership,” he said.
Regardless of the context, namely a forced downsizing internally, at Google and Amazon, Microsoft wants to hammer home its message, according to industry observers.
“Microsoft wants everyone to know that it is heavily involved in cutting-edge AI,” said Nemertes CEO and analyst Johna Till Johnson.
Additionally, Microsoft’s investment in OpenAI also allows the tech giant to take its time developing AI capabilities and not have to be under so much pressure to quickly realize value for its products and services, adds Will McKeon- White, an analyst at Forrester.
So, instead of trying to figure out which areas of AI need the most development or spot the most profitable technologies, Microsoft will let OpenAI lead the research activities, according to the Forrester analyst.
“Being able to rely on a neutral third-party organization, through which all this passes, is a good thing and it makes sense to continue to fund it”, assures Will McKeon-White.
Waiting for Google
With Microsoft’s significant investment in OpenAI, many are wondering how Google will react to this move and the popularity of ChatGPT itself.
Google is reportedly looking for ways to outperform ChatGPT, which could threaten its search design. For the moment, Google’s work consists more of optimizing the results of its search engine, playing on the positioning and highlighting of publications on the Web, according to given themes.
According to the New York Times, Sundar Pichai, CEO of Alphabet, Google’s parent company, recently enlisted Google founders Larry Page and Sergey Brin to help plan and develop a new chatbot and app. image-generating tool, the Times also reports. This would include, among other things, calling on its subsidiary DeepMind and continuing the development of Sparrow, a conversational agent that was the subject of an academic publication at the end of September 2022.
” They [les dirigeants de Google] are acutely aware that their mastery of content search will simply be destroyed if someone else has the ability to deliver high-performance AI [dans ce domaine] said Mr. Johnson.
Google did not immediately respond to a request for comment.
However, Google’s advantage over Microsoft is that it doesn’t need to call on outside resources to develop new AI features, says Johna Till Johnson.
“Historically, Microsoft buys, then closes, and sometimes incorporates companies into its organization,” she says. “Google doesn’t tend to do that. They rely more on home-grown products.”
Google’s biggest problem right now is that many of its generative AI products haven’t launched yet, McKeon-White said.
“They probably have products. It’s just that they’re not quite ready for primetime, they’re more conservative about their availability, or they’re actively integrating it into other systems,” he says.
A larger AI team
Google may also need to beef up its AI development team.
Compared to many other types of technologies, the development of an AI requires “a lot of work”, continues the Forrester analyst.
“OpenAI has been very successful in expanding its talent pool,” he says, adding that the company is otherwise able to devote a large amount of money to a specific problem. On Twitter, Sam Altman, CEO and co-founder of OpenAI praises “the density of talent” within his company, which has 375 employees.
By contrast, Google needs to spread its funds across all of its projects, which span the gamut of AI technologies, according to McKeon-White.
“It’s really the difference between a small organization that is very well funded and very focused and a large company that has to take care of all the activities of a company”, advances the analyst.